ACTM: API Call Transition Matrix-based Malware Detection Method

Abstract

Traditional malware detection techniques, such as signature-based detection and traditional antivirus software, are not beneficial for detecting many recent malware threats. In this paper, we propose a novel malware detection technique, API call transition matrix-based malware detection (ACTM), that efficiently detects malware based on their runtime behavior. We find that the ACTM technique performs better and detects malware with approximately 95.23% accuracy. ACTM can find applications in designing real-time malware detection when an enterprise network security system is concerned.